fix-header dying in _cpp_macroexpand

[Kaveh R. Ghazi]

Zack,

I'm seeing SEGV in stage1 from this morning's CVS when fixproto runs
fix-header.  Here's the backtrace from alphaev5-dec-osf4.0b, but I get
the same on solaris.

 > #0  0x12001c644 in _cpp_macroexpand (pfile=0x11fffe918, hp=0x14002dcc0)
 >     at ../../egcs-CVS20000416/gcc/cpphash.c:986
 > 986       start_line = CPP_BUF_LINE (ip);
 > Breakpoint 1 at 0x3ff800dc994

 > (gdb) where
 > #0  0x12001c644 in _cpp_macroexpand (pfile=0x11fffe918, hp=0x14002dcc0)
 >     at ../../egcs-CVS20000416/gcc/cpphash.c:986
 > #1  0x120011b5c in maybe_macroexpand (pfile=0x11fffe918, written=0)
 >     at ../../egcs-CVS20000416/gcc/cpplex.c:1338
 > #2  0x120011d54 in cpp_get_token (pfile=0x11fffe918)
 >     at ../../egcs-CVS20000416/gcc/cpplex.c:1382
 > #3  0x120006614 in read_scan_file (in_fname=0x11fffef3e
 >     "/usr/var/tmp/gcc-testing/build/gcc/include/stdio.h",
 >     argc=4, argv=0x11fffeb08) at
 >     ../../egcs-CVS20000416/gcc/fix-header.c:664
 > #4  0x120007c30 in main (argc=8, argv=0x11fffeae8)
 >     at ../../egcs-CVS20000416/gcc/fix-header.c:1173

--
Kaveh R. Ghazi			Engagement Manager / Project Services
ghazi@caip.rutgers.edu		Qwest Internet Solutions